Related Experiment Videos
Measurement and classification of retinal vascular tortuosity
1Department of Applied and Numerical Mathematics, Sandia National Laboratories, Alburquerque, NM 87185, USA. wehart@cs.sandia.gov
International Journal of Medical Informatics
|April 8, 1999
Summary
Automated methods accurately measure blood vessel tortuosity in retinal images, aiding ophthalmological diagnostics. These novel tortuosity measures achieve high classification rates for both vessel segments and networks.
Area of Science:
- Ophthalmology
- Medical Imaging
- Computer Vision
Background:
- Ophthalmological diagnostic tools benefit from automated blood vessel tortuosity measurement.
- Quantifying vessel tortuosity is crucial for diagnosing various eye conditions.
Purpose of the Study:
- To develop and evaluate automated tortuosity measures for blood vessel segments from retinal images.
- To assess the effectiveness of these measures in classifying vessel segment and network tortuosity.
Main Methods:
- Extraction of blood vessel segments from RGB retinal images.
- Development of a suite of automated tortuosity measures.
- Evaluation using two classification tasks: segment and network tortuosity classification.
Main Results:
- Achieved 91% classification rate for individual blood vessel segments.
- Achieved 95% classification rate for blood vessel networks.
- Demonstrated that the measures capture ophthalmologists' perception of tortuosity.
Conclusions:
- The developed automated tortuosity measures are effective for ophthalmological applications.
- These measures show high accuracy in classifying tortuosity at both segment and network levels.
- The method of blood vessel segment extraction can influence the accuracy of tortuosity measures.